A modification of the U.S. Geological Survey one-sixth order semiquantitative spectrographic method for the analysis of

U.S. GEOLOGICAL SURVEY CIRCULAR 996 Department of the Interior

TABLES

A MODIFICATION OF THE U.S. GEOLOGICAL SURVEY ONE-SIXTH ORDER SEMIQUANTITATIVE SPECTROGRAPHIC METHOD FOR THE ANALYSIS OF GEOLOGIC MATERIALS THAT IMPROVES LIMITS OF DETERMINATION OF SOME VOLATILE TO MODERATELY VOLATILE ELEMENTS

The semiquantitative emission spectrographic method for the analysis of 30 elements in geologic materials described by Grimes and Marranzino (1968) has been used almost exclusively by the U.S. Geological Survey over the last decade to provide a foundation for the assessment of mineral resource potential of Federal lands. A modification of the method increasing the sensitivity for trace elements provides a broader data base which will allow the geochemist to make more comprehensive geochemical interpretations without supplemental acid dissolution and solvent extractions of samples analyzed by atomic absorption spectroscopy. Data generated using the modification allow the geochemist to make interpretations with greater confidence at low element concentration levels. The use of a compound-pendulum-mounted filter adapted to the existing instrumentation (Jarrell-Ash1 1.5-m optical-emission spectrograph, D.C.- Arc) increases the sensitivity of the method for 12 elements (Ag, As, Au, Be, Bi, Cd, Cr, Cu, Ph, Sb, Sn, and Zn), while the sensitivity of only one element (W) is comprised and the sensitivity for the remaining 17 elements is unchanged. Ideal operating conditions are required to obtain the best results. Therefore, production rates are less than those that can be anticipated using the method described by Grimes and Marranzino, limiting the utility of the adaption to specific research requiring better sensitivity

The modification uses a compound pendulum-mounted filter to regulate the amount of emitted light passing into the spectrograph. A compound pendulum is defined as any friction damped "real" pendulum having a mass distributed over its entire body, a calculable center of gravity.

The pendulum used in the modification has an aluminum arm with an adjustable brass counterweight. The arm is attached to a Teflon hub, and the hub of the pendulum is mounted on a 6-in. rod extendiitg out from an aluminum base which is attached directly onto the arc stand (fig. 1). To maintain frictional consistency over time, a Teflon sleeve is placed on the rod on which the Teflon hub rotates. A 97 o arc of 8-in. radius is attached to the end of the aluminum pendulum in such a way that the arc extends 48.5 o on each side of the attachment point. A l-in. filter is centered on the 0 o position of the arc so that the emitted light path is centered vertically on the filter. A series of three 100-percent transmission ports and four 0-percent transmission spokes extend outward on both sides from the 0 ° position of the arc. This configuration allows for smooth, continuous filtering transition of emitted light. The filter is made from a section of exposed film (Kodak SA-1)1 providing for 40- percent transmission of light above 270.0 nanometers (nm). The pendulum is set in a horizontal position at the start of each burn and held in place by a low-voltage plunger solenoid which is wired to a timer and is automatically activated 20 seconds after the arc is struck, releasing the pendulum. Listed below are the instrumentation and analytical parameters used in the development of the modification.

Spectrograph: 1.5-m Wadsworth-mounted Jarrell-Ash, equipped with a 15,000 lines/in. (590 grooves/mm) lined grating blazed for 310.0 nm providing for a dispersion of 5.45 angstroms/~ over a spectral range of 210.0 nm to 480.0 nm in the second order, slit width of 10 micrometers. Source: Direct-current 220 volts with maximum arc current of 13 amperes Electrodes:

Upper: preformed- 1/8-in. diameter with 8 o

Lower: preformed graphite 1/4-in. diameter, taper (UltraCarbon 5001)

crater wall thickness of 0.04 in., crater diameter of 0.144 in., and crater depth of 0.313 in. (UltraCarbon 7075) Film: 35-mm Spectrum Analysis No. 1 (SA-l) emulsion, processed with D-19 Kodak developer for 2.5 minutes at 20 °C, fixed for 3 minutes, rinsed in cold water and air dried Arc parameters:

Electrode gap: 4 mm Exposure: 120 seconds 0-20 seconds at 100-percent transmission 20-52 seconds transition from 100-percent

52-120 seconds at 40-percent transmission

52-120 seconds at 0-percent transmission for

Current: Initial excitation of 6 amperes increased to 13 amperes immediately

The method allows for 100-percent transmission of light for the initial 20 seconds of the burn. During these initial20 seconds, the majority of the 12 elements (Ag, As, Au, Be, Bi, Cd, Cr, Cu, Ph, Sb, Sn, and Zn) are volatilized. After 20 seconds have elapsed, a solenoid releases the pendulum-mounted filter from a horizontal position, so that it swings freely through the light path with an average period of 0.80 seconds. As the volatile elements are consumed during the 32-second pendulum phase, the distillation of the elements into the arc decreases, and the refactory elements begin to volatilize more rapidly. When the pendulum has stopped, all emitted light is filtered to 40-percent transmittance for the remaining 68 seconds of the burn. During the last phase of the burn, when the 40- percent filter is in place, the majority of the refactory elements are volatilized. Since the free-swinging compound-pendulum bob passes through the optical-emission light path with a decreasing amplitude, it provides for a transitional period of 32 seconds in which the emitted light will be filtered from 100- to 40-percent transmission (fig. 2). Therefore, the period the filter is in the light path increases with time as a function of the decreasing amplitude of the pendulum. The resulting smooth, continuous increase in filtration as expressed by the curve in figure 2 is aided by the configuration of the pendulum bob. Without the transmission ports and spokes on the bob extensions, the curve would be relatively flat over the first 20 seconds and extremely steep over the last 10 seconds of the pendulum phase. transmission to 40 percent transmission

for wavelengths of 270.0 nm and above

wavelengths below 270.0 nm

Aluminum mounting box \olenoid B. -Top View /

The effect of the modification is an increase in sensitivity of the method for the 12 elements: Ag, As, Au, Be, Bi, Ca, Cr, Cu, Ph, Sn, and Zn by a factor of up to 5 (table 1) while the sensitivity of tungsten decreased as a direct response to the filtering process.

Geologic materials have complex and variable matrices. Refractory elements which commonly make up the matrices of these geologic materials cause difficulty in the interpretation of the resulting spectrum by producing line interferences on the volatile-element emission lines. In addition, large

changes in matrix composition of samples can cause changes in line intensities in the spectrum. These variations in composition characteristically cause changes in arc temperatures and the distillation rates of specific elements into the arc. Thus, the change in percent transmittance from 100 to 40 percent must be a gradual transition (fig. 2) rather than a single or multistep transition. This transition helps reduce the effects of matrix variations by providing a relatively long period of time compared to the total burn over which the change in transmittance takes place, and yet, it still produces the desired effects of increasing sensitivity of specific elements while maintaining a good line-tobackground ratio.

w

In order to better understand and effectively eliminate or reduce interference problems, selective volatilization rates (SVRs) have been studied for the 30 elements analyzed routinely in geologic samples. The U.S. Geological Survey Trace Element Glass Standards sample, GSE, was utilized because its uniform concentration range for the major elements is comparable to silicate matrices found in the natural environment.

The SVRs of each of the selected elements in the standard were determined by racking the camera at 10-second intervals during a 180-second excitation period. The 30-element semiquantitative spec- r trographic method (Grimes and Marranzino, 1968) was used to determined the emission concentration for each element at each 10-second interval, and the value was recorded. From these data the 30 elements were grouped into three subsets: volatile, moderately volatile, or refractory The emission concentrations for all elements within each subset were added for each 10-second interval, and the resulting values were then changed to relative percent concentration, which was rounded to the nearest whole percent. Plots of selective volatilization rates for the three subsets were calculated and are shown in figure 3.

The visual determination of line intensity is accurate to within ± 30 percent of the correct value (Ahrens and Taylor, 1961). This accuracy is adequate for the intended purpose of the SVRs in demonstrating general volatilization rates of volatile and retractory elements. We compared visual determination versus densimetric readings for several elements and found the visual determinations were accurate within the limits set forth by Ahrens and Taylor (1961).

The spectrographic conditions under which the SVRs in figure 3 were determined are shown in table 2. However, 100-percent transmission was used for the entire 180-second bum. It was determined from the SVRs that the most effective filtering design is one in which the pendulum phase ends approximately at the intersect point of the volatile and refractory curves.

o - 4%

u are identified with geometric brackets whose these brackets, 1., 0.7, 0.5, 0.3, 0.2, 0.15, 0.1, etc. Precision and accuracy of the pendulum method boundaries are 1.2, 0.83, 0.56, 0.38, 0.26, 0.18, 0.12, etc., but are reported arbitrarily as midpoints of are best demonstrated in table 3, which presents the range of the 10 replicate analyses for each ele- value of that element for the particular reference ment, the mean value, median value, percent rela- material. A general indication of accuracy can be tive standard deviation (RSD), and the accepted obtained by comparing the median and mean values from the replicate analyses to the accepted semiquantitative data that are reported as intervalues. Even though it is not acceptable to make val midpoints, a visual comparison will show a statistical comparison between quantitative and excellent agreement between the experimental Sample Range Median

N(.5) N(.5)

values and the acceptable values. Specific attention is drawn to the test studies of Ag, Bi, Cd, Sb, and Zn in table 4 where the accepted value in a Accepted value Mean RSD

N(.5)

reference sample is at the stated limit of determination and the pendulum method is detecting the elements at these low levels with excellent precision. Sample

GXR-1 1,500-2,000 GXR-2 GXR-3 GXR-4 GXR-5

GXR-3 GXR-4 GXR-5 GXR-6

GSC GSO GSE Range

N(2)-L(2) N{2)

20- N(2)

N(2)

N{2)

N(5)

L(5)- N{5)

N(5) N(5)

N(5)

N(5)

1,800 L(2) N{2)

N(2)

N{2)

N(2)

N(5)

N(5)

N(5) N(5) N(5)

N(5)

N{2)

N(2) N(2)

N(2)

N(5)

N(5) N(5) N(5) N(5)

N(5)

56.0 RSO Accepted value

The precision of the method is also shown by the percent with an average of 9.85 percent. Comrelative standard deviations (RSD) in table 4. The monly acceptable RSDs for the semiquantitative RSDs o~~ained range in magnitude from 0 to 77 emission spectrographic method range from 10 to Range Median

GSB N(5) N(5) N(5) GSC

GSE Accepted value RSD Mean

Cr 30 percent (Mosier, 1972). Of the 10 RSDs greater precision because of the poor background-to-line than 30 percent, 5 can be attributed to analyses ratios on the upper end of the spectrum where the of nickel and cobalt. These two elements show less nickel and cobalt lines that were used are located (Ni, 341.4 nm; Co, 345.3 nm). To help improve the between the grating and the spectrum film, as spectrum at the high end, a glass plate, which pro- close to the film as possible. The plate was posivided an additionallO to 15 percent light filtra- tioned in such a way that it covered the wavetion below 400.0 nm, was placed in the light path lengths of light between 340.0 nm and 350.0 nm. assessment of mineral resource potential in regional geochemical exploration study areas can be im-Semiquantitative analytical data for some proved by utilizing a modification to the one-sixth volatile to moderately volatile elements used in the order semiquantitative emission spectrographic Sample

method described by Grimes and Marranzino to 40 percent over 32 seconds after the initial 20 (1968). The modification uses a compound- seconds of the burn at 100 percent. Filtering more pendulum-mounted filter to provide a transitional light during the latter part of the bum improves change for the filtering of emitted light from 100 the sensitivity of the method for 12 trace elements GXR-1

GXR-3 GXR-4 GXR-5 GXR-6

GXR-1 GXR-2 GXR-3 GXR-4

GSB GSC GSD GSE Range

L( 5)

N(5) N(5) N(5)

N(5) N( 5)

N(5) N(5)

L(5)

N(5) N{5) N(5)

N(5) N(5)

N(5)

N(5)

N(5)

L{5)

N(5) N(5)

N(5)

N(5) N(5)

N(5)

N(5)

N(5) RSD Accepted value

v

commonly used in mineral exploration by reduc- causing interferences without compromising preciing spectral background and the line emission of sion and accuracy. Using this modification greatly elements commonly responsible for problem- reduces the capacity of the method in terms of

w

Limit of determination (ppm) Element Lines used (nm)

Au B Ba Be

Cu Fe La Mg

sample throughput. However, this study should provide a basis for the design of an automated mechanism allowing for greater production. Lines used (nm)

v

Where this page came from

This page was imported from U.S. Geological Survey. Published by the U.S. Geological Survey and, as a work of the United States government, in the public domain.

Nobody has written it yet — it is the source material at a new address, which is why search engines are asked to skip it and why no one earns from it. It is up for grabs: take it on, and it is yours to rewrite and to earn from.